1. Field of the Invention
The present invention relates to a quality control apparatus and method using a bar code data entry system. More particularly, this invention relates to such an apparatus and method in which production data from all phases of the production process, from manufacturing through shipment of finished production units, are entered into a production information system through a bar code data entry system. The production information system stores production history data for each individual production unit as it progresses through the production process. The production information system also analyzes this data on a real-time basis to identify the existence and sources of production problems, such as recurrent defects arising from errors in the production process. Real-time analysis of production history data for individual production units enables the production facility's process engineering team to prevent shipment of defective units and to increase production efficiency by identifying and eliminating the sources of recurrent defects before numerous defective units have been produced. The present invention therefore improves product quality and production efficiency through the collection, processing, and analysis of accurate production information in real time.
2. Discussion of the Related Art
Successful manufacturing concerns have realized that they must apply their utmost energy and ingenuity to the problems of improving product quality and production efficiency if they expect to compete in the contemporary commodities markets. In the technology sector, in particular, these markets are characterized by fierce competition among producers and increasingly sophisticated expectations by consumers. Survival in this brisk environment demands that manufacturers simultaneously pursue development of advanced technology, minimize every controllable cost factor, and create new products and product features to anticipate evolving consumer preferences. Integrated manufacturing information systems have become essential tools for firms to successfully meet these challenges.
Advanced production information systems process and communicate large amounts of information of widely diverse types, often collected at a variety of locations and from numerous sources, with the objective of providing practical solutions to critical production issues. But a serial paradigm for data collection, defect analysis, and solution implementation cannot provide the proactive methodologies required to optimize dynamic manufacturing environments and thereby to maintain a competitive position in the contemporary marketplace. The reason is evident: effective production analysis requires compilation of production histories for at least a representative sample of individual production units. Also, each such history must include complete and accurate data on the specific defects found in the production unit and when those defects were first detected. But when production defects are identified in individual units of a product only after the units have been completed and placed in service, it becomes exceedingly difficult to analyze the defects and trace them back to their probable sources at the various stages of the production process.
Real-time collection and analysis of production data has become a central issue in the design of modem industrial information systems. On the other hand, real-time data collection is particularly difficult to carry out in the context of general manufacturing operations. This difficulty arises from the requirement for effective real-time analysis that data collection be rapid and substantially error-free. Conventional quality control systems, in contrast, have relied upon manual recordation and encoding of production data prior to analysis by the information system. Manual entry of data into the computer's database degrades the effectiveness of automated production analysis for two basic reasons. First, data thus entered unavoidably contains encoding errors, whether from miskeying of the data codes or from misinterpretation or simple loss of the information to be encoded. Second, manual data entry creates an inherent processing bottleneck in the analysis process because human operators cannot achieve data entry rates even approaching those of automated encoding systems. These dual aspects of the data entry problem are exacerbated when the personnel performing the data entry also have primary responsibility for generating the information to be encoded, as is the case in efficient test and inspection situations.
Automated data input, to accelerate data entry and prevent data loss and encoding errors, therefore offers a potentially important key to solving the problem of real-time production analysis. Various approaches have been proposed to automate production data entry, including using computer keyboards, optical card readers, and magnetic cards. The basic weakness of manual data entry persists in these methods, though, because they all depend upon human operators to translate specific facts from a human-intelligible form into machine-readable code. That is, these existing data entry approaches still carry the disadvantage of requiring human operators to generate computer-readable codes representative of the information to be entered. Voice recognition systems have recently been suggested as a potential alternative, but their effectiveness has not yet been demonstrated.
Effective automated data entry methodologies therefore constitute a primary objective for the next generation of manufacturing information systems. A radical approach, which has yielded some benefit in a limited range of specialized application areas, focuses on eliminating human error by eliminating human involvement in the data collection process. For the present and the foreseeable future, however, no automated system can approach the capacity of human technicians to continuously adapt their operations to complex and rapidly evolving manufacturing environments and to draw subtle but important distinctions between similar fact situations presented by those environments. To eliminate human involvement in data collection is therefore to sacrifice the unique advantages provided by human perception to identify and distinguish defect conditions.
A variety of systems have been proposed to integrate more effectively production personnel and production information systems. U.S. Pat. Nos. 5,088,045 and 5,586,038, for example, both disclose production management systems that include scanning bar coded identification labels to input production unit serial numbers or model numbers into the system's computer. But the former relies on automatic sensors to detect and report defects and relegates personnel to entering binary status data on rework jobs through touch screen terminals. In the latter system, assembly status data is registered by scanning a bar code formed when two halves of an electrical connector are correctly joined. This technique, while undeniably ingenious, has an obviously limited range of applicability; and more importantly, neither of these systems make effective use of the perceptual capacities of the human operators involved in the manufacturing process.
Developments in an alternative direction have attempted to emphasize the advantages of human perception in testing and inspection operations. U.S. Pat. No. 5,086,397, for example, shows a system for collecting data through test and inspection stations where technicians identify and classify manufacturing defects and then enter appropriate data using optical styluses at display terminals. The entry process, however, requires the technician to advance incrementally through a hierarchy of classification menus and to encode substantial amounts of defect classification data through a 10-key numeric pad displayed on the terminal screen. Each defect datum entry therefore requires several manual steps and again depends on the accuracy of manual encoding. This system also does generate not production history information for individual production units, either, because it does not employ serial numbers to identify individual production units.
What the next generation of production information systems urgently needs, and what even the sophisticated systems discussed above do not provide, is a real-time data collection system that takes full advantage of data entry automation while not sacrificing the information quality afforded by human perception. Such a system should completely eliminate manual encoding of defect data while including human technicians to actually generate the data. Preferably, it would rely on existing information and data entry technologies and could be implemented in existing manufacturing facilities with only modest start-up expenses. Ideally, this system would provide reliable and flexible data collection with only minimal operating costs.